Faced with the job of specifying and setting up a measurement system, many engineers may not know that hidden costs can delay or derail the project. Those hidden costs come as direct expenses (money) and resources (time). In many cases, engineers have more control over the latter than they do over a project's budget. Given a cost of about $100/hour for an engineer's time, time saved can amount to a considerable sum.
Where in a project can savings arise, and how do engineers realize them? Most new test systems go through five phases:
And engineers can extract savings from almost all of these activities. The suggestions below can help obtain them.
Specifying a system requires gathering information about measurement requirements and obtaining information about hardware and software that will meet those needs. Save time by ensuring a project has clear, written goals that upper managers have approved. A clear objective lets team members work together efficiently to reach a goal. Without clear objectives, people waste time trying to decide what to do.
When it comes time to evaluate hardware and software, use suppliers' websites to obtain current product information. Vendors often provide "configurators" that translate engineers' measurement-system requirements into suggested product lists or bills of materials. These online tools help ensure the selection of compatible hardware and software. Save time by using configurators' listings to compare capabilities and characteristics of products from several vendors.
Buying hardware and software involves mainly spending money, so it's unlikely you'll save much time during this activity. But the hidden costs of delayed deliveries, out of stock products, and poor technical support will drain time from other jobs. Do your best to ensure vendors can meet your schedules.
Setup and configuration of a test system can take much time, so engineers can look for savings in this activity. Software tools supplied by hardware vendors should let users quickly establish that all the functions on an add-in data-acquisition board, for example, work properly. These tools may provide routines that display data, control I/O ports, or perform other simple tests. Configuration utilities should offer soft front panels—images on a monitor that look like instruments and let users see and control operations.
In some cases, these utilities offer signal-conditioning capabilities, such as conversion of thermocouple voltages into meaningful temperatures. Sophisticated configuration software can produce instrument-control routines for use in programs written in Visual Basic, C, or LabVIEW. This sort of automated setup saves time. These days, engineers should not settle for low-level drivers that simply "connect" an instrument or add-in card with application code.
If possible, engineers should ensure the drivers they use in a project won't become obsolete with the next version or upgrade of a product, such as a data-acquisition or I/O board. Without this assurance, established test-system software may require extensive revision when test systems get updated or duplicated using new equipment. The software functions within drivers should remain "constant" from one generation of product to the next. (The driver software may change, but such changes should remain transparent to users.)
The proper choice of hardware also can save configuration time. Newer sensors, for example, come with embedded devices that carry specifications and identifying information. The Transducer Electronic Datasheet (TEDS) format lets data-acquisition systems determine the sensors attached to them so they can alert software to specific sensor types, calibration needs, linearization requirements, and so on. In many cases, smart-sensor hookups require little time, which translates into time savings, particularly for systems that include hundreds of sensors.
Application development requires considerable time, too. After all, engineers need time to write—and test—code that will acquire data, establish closed-loop controls, display test results, communicate over a network, and so on. But instead of starting from scratch, first investigate libraries of sample programs. For the most part, sample programs won't just drop into an application, but they can show how others solved similar problems. Short code "snippets" as well as tested algorithms and macros can save considerable time.
User groups and news groups on the Internet also provide information that may be used in order to reduce development time. (These sources also may offer product reviews and comparison information that can guide decisions during a system's specification.) And vendor sites include white papers and application notes that describe how users tackled measurement projects. Recent articles on magazines' websites also contain practical information about implementing measurement systems.
Because validation and calibration takes a small portion of a project's time and money budget, it doesn't offer many ways to economize. But taking advantage of configuration software, using smart sensors, and obtaining help from online white papers and code libraries should simplify and shorten validation and calibration tasks.
All of the time and money savings mentioned above involve tradeoffs, many of which involve difficult choices. Just remember, though, more often than not it's much easier to get a few more dollars for a project than to ask for an additional four weeks to complete it.